3 ms·
I understand that surveys are designed to have a random sample whose findings can be generalized to the population. But even so, polling ~1000 people and apply
by rchaud 3y ago
I understand that surveys are designed to have a random sample whose findings can be generalized to the population.
But even so, polling ~1000 people and applying those opinions as generalizable to the US where 170 million use the app doesn't sound right to me.
Especially when the poll took place at a time when the news cycle was focused on parroting the government's case for banning the app.
I think TT is dangerous in much the same way that FB/IG/TW is: disinformation and division is rampant, and the app owners are incentivized to propagate it because of engagement. That's intrinsic to social media, as we saw in the 2016 and 2020 elections. Singling out TT as uniquely guilty in this regard doesn't help protect Americans from the influence operations happening on the other apps. You can decide for yourself if what is posted on FB groups or Truth Social is doing a better job of uniting Americans.
- alephnerd 3y ago> But even so, polling ~1000 people and applying those opinions as generalizable to the US where 170 million use the app doesn't sound right to me. The Central Limit Theorem [0] is why a poll of ~1k is fairly representative. Everybody keeps complaining about the size of a poll while ignoring grade school math (not aimed at you - I've seen this talking point repeated ad nauseum). [0] - https://sphweb.bumc.bu.edu/otlt/mph-modules/bs/bs704_probability/BS704_Probability12.html https://sphweb.bumc.bu.edu/otlt/mph-modules/bs/bs704_probabi...
- squigz 3y agoWhy don't we do this for elections?
- amenhotep 3y agoBecause people wouldn't like it.
- alephnerd 3y agoThey do use the CLT in polls for elections. There are different types of polls (some have crap methodology but others are actually statistically driven).
- gus_massa 3y agoPeople lie, sometimes they want to vote someone but are afraid and tell the poller another candicate. The electtion system in USA is very sensitive to small differences, that are below the extimated error of the pool. ("swing states") The looser will always ask for a full election.